The emergence and spread of drug-resistant Plasmodium falciparum poses an immense global health threat. While we understand much about the selection of drug resistant malaria in populations, little is known about in- host evolution. Most individuals infected with falciparum malaria carry multiple genetically distinct variants (genotypes, strains) which continually evolve and compete for resources. This within-host competition could be as important as competition within populations. Many of the variants are present at low levels and undetectable by older genotyping technologies. In order to measure within-host competition, we propose to employ a new method uniquely capable of identifying and quantitating genotypes in a single host - Massively Parallel Pyrosequencing (MPP). With this technology, we will measure, within individual hosts, the true diversity of falciparum infections and the selective pressure of antimalarials. Specifically, we will: i) sequence and quantitate merozoite surface protein-2 (msp2) genotypes within individual subjects from two areas of different transmission intensity: Tanzania (high transmission) and Thailand (low intensity), ii) measure the rate of change in frequency (selection coefficients) for individual parasite variants from Tanzanian patients treated with Coartem and determine the effect of competing variants on frequency, and iii) measure up-selection for individual parasite variants from Thai patients treated with Artesunate-Mefloquine to define selection coefficients and better define the phenotype used to study drug resistance to artemisinins. This novel approach of quantitatively studying variant diversity within individuals will i) allow accurate measurement of within-host selection and dynamics of variants in mixed infections, ii) provide new tools for testing hypotheses about the genetic basis of low- and high-level resistance, and iii) define the factors that allow the development and spread drug resistant parasites in populations.